iommufd: Prepare for IOMMUFD_OBJ_HW_QUEUE

[PATCH v2 05/14] iommufd: Return EOPNOTSUPP for failures due to driver bugs

Posted by Nicolin Chen 3 months, 4 weeks ago

It's more accurate to report EOPNOTSUPP when an ioctl failed due to driver
bug, since there is nothing wrong with the user space side.

Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
---
 drivers/iommu/iommufd/device.c       | 2 +-
 drivers/iommu/iommufd/hw_pagetable.c | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/iommu/iommufd/device.c b/drivers/iommu/iommufd/device.c
index ed0dc539d490..e9b6ca47095c 100644
--- a/drivers/iommu/iommufd/device.c
+++ b/drivers/iommu/iommufd/device.c
@@ -1483,7 +1483,7 @@ int iommufd_get_hw_info(struct iommufd_ucmd *ucmd)
 		 */
 		if (WARN_ON_ONCE(cmd->out_data_type ==
 				 IOMMU_HW_INFO_TYPE_NONE)) {
-			rc = -ENODEV;
+			rc = -EOPNOTSUPP;
 			goto out_free;
 		}
 	} else {
diff --git a/drivers/iommu/iommufd/hw_pagetable.c b/drivers/iommu/iommufd/hw_pagetable.c
index 8565a6f596b2..fe789c2dc0c9 100644
--- a/drivers/iommu/iommufd/hw_pagetable.c
+++ b/drivers/iommu/iommufd/hw_pagetable.c
@@ -264,7 +264,7 @@ iommufd_hwpt_nested_alloc(struct iommufd_ctx *ictx,
 	hwpt->domain->cookie_type = IOMMU_COOKIE_IOMMUFD;
 
 	if (WARN_ON_ONCE(hwpt->domain->type != IOMMU_DOMAIN_NESTED)) {
-		rc = -EINVAL;
+		rc = -EOPNOTSUPP;
 		goto out_abort;
 	}
 	return hwpt_nested;
@@ -321,7 +321,7 @@ iommufd_viommu_alloc_hwpt_nested(struct iommufd_viommu *viommu, u32 flags,
 	hwpt->domain->cookie_type = IOMMU_COOKIE_IOMMUFD;
 
 	if (WARN_ON_ONCE(hwpt->domain->type != IOMMU_DOMAIN_NESTED)) {
-		rc = -EINVAL;
+		rc = -EOPNOTSUPP;
 		goto out_abort;
 	}
 	return hwpt_nested;
-- 
2.43.0

RE: [PATCH v2 05/14] iommufd: Return EOPNOTSUPP for failures due to driver bugs

Posted by Tian, Kevin 3 months, 3 weeks ago

> From: Nicolin Chen <nicolinc@nvidia.com>
> Sent: Saturday, June 14, 2025 2:35 PM
> 
> It's more accurate to report EOPNOTSUPP when an ioctl failed due to driver
> bug, since there is nothing wrong with the user space side.
> 
> Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>

Reviewed-by: Kevin Tian <kevin.tian@intel.com>

Re: [PATCH v2 05/14] iommufd: Return EOPNOTSUPP for failures due to driver bugs

Posted by Pranjal Shrivastava 3 months, 3 weeks ago

On Fri, Jun 13, 2025 at 11:35:17PM -0700, Nicolin Chen wrote:
> It's more accurate to report EOPNOTSUPP when an ioctl failed due to driver
> bug, since there is nothing wrong with the user space side.
> 
> Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
> ---
>  drivers/iommu/iommufd/device.c       | 2 +-
>  drivers/iommu/iommufd/hw_pagetable.c | 4 ++--
>  2 files changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/iommu/iommufd/device.c b/drivers/iommu/iommufd/device.c
> index ed0dc539d490..e9b6ca47095c 100644
> --- a/drivers/iommu/iommufd/device.c
> +++ b/drivers/iommu/iommufd/device.c
> @@ -1483,7 +1483,7 @@ int iommufd_get_hw_info(struct iommufd_ucmd *ucmd)
>  		 */
>  		if (WARN_ON_ONCE(cmd->out_data_type ==
>  				 IOMMU_HW_INFO_TYPE_NONE)) {
> -			rc = -ENODEV;
> +			rc = -EOPNOTSUPP;
>  			goto out_free;
>  		}
>  	} else {
> diff --git a/drivers/iommu/iommufd/hw_pagetable.c b/drivers/iommu/iommufd/hw_pagetable.c
> index 8565a6f596b2..fe789c2dc0c9 100644
> --- a/drivers/iommu/iommufd/hw_pagetable.c
> +++ b/drivers/iommu/iommufd/hw_pagetable.c
> @@ -264,7 +264,7 @@ iommufd_hwpt_nested_alloc(struct iommufd_ctx *ictx,
>  	hwpt->domain->cookie_type = IOMMU_COOKIE_IOMMUFD;
>  
>  	if (WARN_ON_ONCE(hwpt->domain->type != IOMMU_DOMAIN_NESTED)) {
> -		rc = -EINVAL;
> +		rc = -EOPNOTSUPP;
>  		goto out_abort;
>  	}
>  	return hwpt_nested;
> @@ -321,7 +321,7 @@ iommufd_viommu_alloc_hwpt_nested(struct iommufd_viommu *viommu, u32 flags,
>  	hwpt->domain->cookie_type = IOMMU_COOKIE_IOMMUFD;
>  
>  	if (WARN_ON_ONCE(hwpt->domain->type != IOMMU_DOMAIN_NESTED)) {
> -		rc = -EINVAL;
> +		rc = -EOPNOTSUPP;
>  		goto out_abort;
>  	}
>  	return hwpt_nested;

Much clearer retvals.
Reviewed-by: Pranjal Shrivastava <praan@google.com>

> -- 
> 2.43.0
>

Re: [PATCH v2 05/14] iommufd: Return EOPNOTSUPP for failures due to driver bugs

Posted by Jason Gunthorpe 3 months, 3 weeks ago

On Fri, Jun 13, 2025 at 11:35:17PM -0700, Nicolin Chen wrote:
> It's more accurate to report EOPNOTSUPP when an ioctl failed due to driver
> bug, since there is nothing wrong with the user space side.
> 
> Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
> ---
>  drivers/iommu/iommufd/device.c       | 2 +-
>  drivers/iommu/iommufd/hw_pagetable.c | 4 ++--
>  2 files changed, 3 insertions(+), 3 deletions(-)

Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>

Jason

Re: [PATCH v2 05/14] iommufd: Return EOPNOTSUPP for failures due to driver bugs

Posted by Baolu Lu 3 months, 3 weeks ago

On 6/14/25 14:35, Nicolin Chen wrote:
> It's more accurate to report EOPNOTSUPP when an ioctl failed due to driver
> bug, since there is nothing wrong with the user space side.
> 
> Signed-off-by: Nicolin Chen<nicolinc@nvidia.com>

Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>